/usr/share/pyshared/xmldiff/mydifflib.py is in xmldiff 0.6.10-2build1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 | """
longest common subsequence algorithm
the algorithm is describe in "An O(ND) Difference Algorithm and its Variation"
by Eugene W. MYERS
As opposed to the algorithm in difflib.py, this one doesn't require hashable
elements
"""
def lcs2(X, Y, equal):
"""
apply the greedy lcs/ses algorithm between X and Y sequence
(should be any Python's sequence)
equal is a function to compare X and Y which must return 0 if
X and Y are different, 1 if they are identical
return a list of matched pairs in tuplesthe greedy lcs/ses algorithm
"""
N, M = len(X), len(Y)
if not X or not Y :
return []
max = N + M
v = [0 for i in xrange(2*max+1)]
common = [[] for i in xrange(2*max+1)]
for D in xrange(max+1):
for k in xrange(-D, D+1, 2):
if k == -D or k != D and v[k-1] < v[k+1]:
x = v[k+1]
common[k] = common[k+1][:]
else:
x = v[k-1] + 1
common[k] = common[k-1][:]
y = x - k
while x < N and y < M and equal(X[x], Y[y]):
common[k].append((x, y))
x += 1 ; y += 1
v[k] = x
if x >= N and y >= M:
return [ (X[x],Y[y]) for x,y in common[k] ]
def lcs4(X, Y, equal):
"""
apply the greedy lcs/ses algorithm between X and Y sequence
(should be any Python's sequence)
equal is a function to compare X and Y which must return 0 if
X and Y are different, 1 if they are identical
return a list of matched pairs in tuplesthe greedy lcs/ses algorithm
"""
N, M = len(X), len(Y)
if not X or not Y :
return []
max = N + M
v = [0 for i in xrange(2*max+1)]
vl = [v]
for D in xrange(max+1):
for k in xrange(-D, D+1, 2):
if k == -D or k != D and v[k-1] < v[k+1]:
x = v[k+1]
else:
x = v[k-1] + 1
y = x - k
while x < N and y < M and equal(X[x], Y[y]):
x += 1 ; y += 1
v[k] = x
if x >= N and y >= M:
# reconstruction du chemin
vl.append(v)
vl_saved = vl[:]
path = [ ]
k = N-M
while vl:
oldv = vl.pop(-1)
oldk = k
if k==-D or k!= D and oldv[k-1]<oldv[k+1]:
xs = oldv[k+1]
k = k + 1
else:
xs = oldv[k-1]+1
k = k - 1
#print "-> x=%d y=%d v=%r ok=%d k=%d xs=%d D=%d" % (x,y,oldv,oldk,k,xs,D)
while x>xs:
x -= 1
y -= 1
#print "(%d,%d)" % (x,y)
path.append( (X[x],Y[y]) )
D -= 1
x = oldv[k]
y = x - k
#print "<- x=%d y=%d v=%r ok=%d k=%d xs=%d D=%d" % (x,y,oldv,oldk,k,xs,D)
#print x,y,deltax,deltay,oldv, oldk, k
path.reverse()
return path #, vl_saved
vl.append(v[:])
def pprint_V( V, N, M ):
for v in V:
for k in xrange(-N-M,N+M+1):
print "% 3d" % v[k],
print
def lcs3( X, Y, equal ):
N = len(X)+1
M = len(Y)+1
if not X or not Y :
return []
# D(i,j) is the length of longest subsequence for X[:i], Y[:j]
pre = [0]*M
row = [0]*M
B = [ [0]*M for i in xrange(N) ]
for i in xrange(1,N):
for j in xrange(1,M):
if equal(X[i-1],Y[j-1]):
row[j] = pre[j-1] + 1
B[i][j] = 2 # move back (-1,-1)
elif pre[j] >= row[j-1]:
row[j] = pre[j]
B[i][j] = 1 # move back (0,-1)
else:
row[j] = row[j-1]
B[i][j] = 0 # move back (-1,0)
pre,row=row,pre
i = N - 1
j = M - 1
L = []
while i>=0 and j>=0:
d = B[i][j]
#print i,j,d
if d == 0:
j -= 1
elif d == 1:
i -= 1
else:
i -= 1
j -= 1
L.append( (X[i], Y[j]) )
L.reverse()
#from pprint import pprint
#pprint(D)
#pprint(B)
return L
try:
import maplookup
lcs2 = maplookup.lcs2
#lcs2 = lcs4
except:
pass
def lcsl(X, Y, equal):
"""return the length of the result sent by lcs2"""
return len(lcs2(X,Y,equal))
def quick_ratio(a,b):
"""
optimized version of the standard difflib.py quick_ration
(without junk and class)
Return an upper bound on ratio() relatively quickly.
"""
# viewing a and b as multisets, set matches to the cardinality
# of their intersection; this counts the number of matches
# without regard to order, so is clearly an upper bound
if not a and not b:
return 1
fullbcount = {}
for elt in b:
fullbcount[elt] = fullbcount.get(elt, 0) + 1
# avail[x] is the number of times x appears in 'b' less the
# number of times we've seen it in 'a' so far ... kinda
avail = {}
availhas, matches = avail.has_key, 0
for elt in a:
if availhas(elt):
numb = avail[elt]
else:
numb = fullbcount.get(elt, 0)
avail[elt] = numb - 1
if numb > 0:
matches = matches + 1
return 2.0 * matches / (len(a) + len(b))
try:
import os
if os.environ.get('PYLINT_IMPORT') != '1': # avoid erros with pylint
import psyco
psyco.bind(lcs2)
except Exception, e:
pass
def test(lcs2=lcs2):
"""
FIXME this should go into the test suite.
"""
import time
t = time.clock()
quick_ratio('abcdefghijklmnopqrst'*100, 'abcdefghijklmnopqrst'*100)
print 'quick ratio :',time.clock()-t
lcs2('abcdefghijklmnopqrst'*100, 'abcdefghijklmnopqrst'*100,
lambda x, y : x==y)
print 'lcs2 : ',time.clock()-t
quick_ratio('abcdefghijklmno'*100, 'zyxwvutsrqp'*100)
print 'quick ratio :',time.clock()-t
lcs2('abcdefghijklmno'*100, 'zyxwvutsrqp'*100, lambda x, y : x==y)
print 'lcs2 : ',time.clock()-t
quick_ratio('abcdefghijklmnopqrst'*100, 'abcdefghijklmnopqrst'*100)
print 'quick ratio :',time.clock()-t
lcs2('abcdefghijklmnopqrst'*100, 'abcdefghijklmnopqrst'*100,
lambda x, y : x==y)
print 'lcs2 : ',time.clock()-t
quick_ratio('abcdefghijklmno'*100, 'zyxwvutsrqp'*100)
print 'quick ratio :',time.clock()-t
lcs2('abcdefghijklmno'*100, 'zyxwvutsrqp'*100, lambda x, y : x==y)
print 'lcs2 : ',time.clock()-t
def main( lcs2=lcs2 ):
print "abcde - bydc"
print lcsl('abcde', 'bydc', lambda x, y : x==y)
for a in lcs2('abcde', 'bydc', lambda x, y : x==y):
print a
print "abacdge - bcdg"
print lcsl('abacdge', 'bcdg', lambda x, y : x==y)
for a in lcs2('abacdge', 'bcdg', lambda x, y : x==y):
print a
import random
def randstr( lmin, lmax, alphabet ):
L = random.randint( lmin, lmax)
S = []
N = len(alphabet)-1
for i in range(L):
S.append( alphabet[random.randint(0,N)] )
return "".join(S)
def randtest():
"""Generate random test sequences and compare lcs2, lcs3, lcs4"""
_cmp = lambda x,y:x==y
import maplookup
lcsm = maplookup.lcs2
_alpha = "abcdefghijklmnopqrstuvwxyz"
while 1:
S1 = randstr( 2,5,_alpha )
S2 = randstr( 2,5,_alpha )
print S1, S2
R1 = lcs2( S1, S2, _cmp )
print "lcs2:", "".join( [ x[0] for x in R1 ] )
R2 = lcs4( S1, S2, _cmp )
print "lcs4", "".join( [ x[0] for x in R2 ] )
R3 = lcsm( S1, S2, _cmp )
print "lcsm", "".join( [ x[0] for x in R3 ] )
print
assert R1==R2
assert R1==R3
if __name__ == '__main__':
main()
|